This weekly update highlights major advances in AI agents – computer programs that can make decisions and do tasks on their own. The big story is that the Linux Foundation created a new group called the Agentic AI Foundation on December 9th to help different companies work together on AI agent technology. Three important tools are joining this foundation: Model Context Protocol (MCP) from Anthropic, goose from Block, and AGENTS.md from OpenAI.

Microsoft shared exciting news about Agent Lightning, a new way to teach AI agents to learn better from their mistakes. This technology helps the agents improve their answers without needing to rewrite all the code. Meanwhile, IBM Research unveiled Project ALICE, which uses multiple AI agents working together to find computer bugs faster and fix them.

Companies are moving quickly to use AI agents in real work. Gartner predicts that by 2028, AI agents will make 15% of work decisions automatically, up from zero in 2024. Construction companies are already using AI agents to plan and track building projects, and tools like Cursor, Make.com, and Zapier are making it easier for regular people to build their own AI agents without being expert programmers.

Experts point out there are still challenges to solve. Companies need to handle data security, make sure agents understand situations correctly, and keep humans involved in important decisions. As these technologies improve, AI agents are becoming the next big thing in artificial intelligence, moving beyond simple chatbots to systems that can actually accomplish goals and work with other AI agents.